Chromatic aberration puzzle

[StewartR]

One of our customers hired a camera recently to take some photos at an air show. His lens, our camera. He's evaluating the camera as a possible purchase. However he's not sure about it because he's finding that he sometimes gets problems with chromatic aberration that he hasn't seen before and doesn't experience with his camera.

Here's a sample shot. It's a very small portion of the leading edge of an aircraft wing, positioned towards the bottom of the frame in the image, presented here at 200% magnification for clarity. I've created four versions:

1. the original (straight out of camera with lens profile corrections applied in Lightroom);
2. with the CA corrected by Lightroom
3. with the saturation boosted to 100% in Lightroom, to highlight where the CA is occurring
4. with the saturation boosted and CA corrected, to highlight any residual CA

All in all this is probably about what I'd expect. Lightroom has done a pretty good job on the CA, though there are some small traces of it left it you look hard (eg by zooming to 200% and increasing the saturation!)

But now here's another portion of the same image. This is the opposite wingtip of the plane, positioned towards the top of the frame in the image. Again it's presented here at 200% magnification, and with the same four versions.

You can see in the original and in the saturated version that the CA is broadly similar to the first portion of the image; perhaps a little worse, but not much. But look at the two versions where Lightroom claims to have corrected the CA - it's noticeably worse, not better!

So anyone help work out what's happening? Camera fault? Lens fault? Lightroom fault? Or what?

 

[boyfalldown]

I'm guessing here. As you know CA occurs when a lens is unable to focus all the colour wavelength on to the same plane (forgive the pun).

I guess that cause the wings are likely on different focal planes the fringe colours in the CA are slightly different as well. Is he using an automatic or a manual correction in LR?

 

[Terrywoodenpic]

I do not see how it can be the cameras fault as chromatic aberration is created by the lens. the camera simply records the image.
In this instance the camera has recorded it and the lens faults very well.

I do not understand why lightroom would make matters worse, in the second image it is doing a terrible job. and not good in the first.

I would expect lightroom to remove chromatic aberration pretty much completely with recent lens profiles.

 

[StewartR]

I do not see how it can be the cameras fault as chromatic aberration is created by the lens. the camera simply records the image.
In this instance the camera has recorded it and the lens faults very well.

I think it might not be as simple as that though. Chromatic aberration is caused by different wavelengths of light bring focused at different distances. Usually one assumes that the sensor is perpendicular to the axis of the lens, so light bring focused at different distances causes the fringing.

But suppose we imagine for a moment that the camera has a fault in that the sensor is slightly out of alignment - say too far forward at the bottom and too far back at the top - so it's not perpendicular to the axis of the lens. What happens then? Do you get CA in one direction but not in another direction, or what? I can't work it out.

 

[StewartR]

I do not understand why lightroom would make matters worse, in the second image it is doing a terrible job. and not good in the first.

I would expect lightroom to remove chromatic aberration pretty much completely with recent lens profiles.

My thoughts exactly.

 

[StewartR]

I guess that cause the wings are likely on different focal planes the fringe colours in the CA are slightly different as well.

The plane is banked steeply in the photo, but even so the nearer wingtip (the lower one, my first example) is going to be a bit closer to the camera than the upper wingtip. I haven't yet tried to calculate the depth of field, but that could be done if it's relevant and important. (I suspect it isn't.)

But, visually, the CA on both wings looks fairly similar before 'correction'. I can't work out why Lightroom would make one better and one worse.

Is he using an automatic or a manual correction in LR?

He isnt. I am. I have a copy of the raw file. I used the automatic CA correction.

 

[dcash29]

I think the CA fix on the first image has been overdone

What camera, what lens?

 

[kendo1]

The camera can be checked by using another lens.
There's a possibility that the lens is decentred, more likely than the camera sensor being out.

 

[frod]

maybe open a support ticket with Adobe?

 

[boyfalldown]

He isnt. I am. I have a copy of the raw file. I used the automatic CA correction.

I wonder if correcting manually would do it. But that doesn't solve why initially

Is a higher resolution camera showing a weakness in the lens

 

[StewartR]

I think the CA fix on the first image has been overdone

Remember though, these aren't two separate images; they're two parts of the same image. Presumably Adobe is looking at the whole image and calculating some sort of global shift to get the best average CA correction across the whole image, and that might be a bit too much for some areas and not quite enough for others.

What camera, what lens?

Nikon D500, Nikon AF-S 80-400mm f/4.5-5.6 G ED VR. Is that important?

 

[ancient_mariner]

Somehow the correction is going too far, and is actually reversing the colours - blue to orange & orange to blue. I can't say why the auto correction is failing, but it should be possible to correct manually, possibly using the dropper on the wingtip.

 

[dcash29]

Surely the 'customer' has taken the same image using his own camera with the lens as a comparison?

 

[Terrywoodenpic]

On my little Fuji X30 all aberrations are corrected in firmware. I have not found any chromatic aberration at all... though the lens is almost certainly producing some.
It seems that Photoshop raw processor has the correct algorithms to correct these things for this lens automatically.

I would suggest that his lens has at least one of the elements decentered to get crossed chromatic aberration like that.

The question of a lens or sensor being mounted out of true, this is used in large format and tilt lenses with out this being any sort of additional problem. And certainly does not add chromatic problems.